The 3GPP (3rd Generation. Partnership Project Radio Access Network) has established LTE (Long Term Evolution) Rel. 8 (Release 8) and the extended version of LTE, which is LTE Rel. 10 (LTE-Advanced). In these standards, a base station, and a radio communication terminal (also called “UE (User Equipment)” and referred to below as a terminal) transmit control information for transmitting and receiving data using a downlink PDCCH (physical downlink control channel) (refer to Non-Patent Literature 1 to 3). FIG. 1 shows the subframe configuration of the downlink. In the subframes, the PDCCH that transmits a control signal and the PDSCH (physical downlink shared channel) that transmits a data signal are time-division multiplexed. The terminal first decodes the control information transmitted to the terminal through the PDCCH and obtains information regarding a frequency allocation required for data reception on the downlink, and adaptive control, for example. The terminal then decodes data for the terminal that is included in the PDSCH, based on the control information. If control information that permits data transmission in the uplink is included in the PDCCH, the terminal transmits data on the PUSCH (physical uplink shared channel) of the uplink, based on the control information.
In order to transmit and receive data on the downlink, an HARQ (hybrid automatic request) combining error correction decoding and an automatic retransmission request has been introduced. After performing error correction decoding, the terminal determines whether or not the data is correctly decoded, based on a CRC (cyclic redundancy checksum) added to the data. If the decoding is successful, the terminal feeds back an ACK to the base station. If, however, the decoding fails, the terminal feeds back a NACK to the base station, prompting retransmission of the data in which an error is detected. The feedback of ACK/NACK (acknowledge response; hereinafter referred to as “A/N”) is transmitted in the uplink. If data is not assigned to the PUSCH at the time of transmission, transmission is performed on PUCCH (physical uplink control channel). If, however, data is assigned to the PUSCH at the time of A/N transmission, A/N is transmitted on either PUCCH or PUSCH. When this is done, the base station instructs the terminal beforehand as to whether transmission is to be done on PUCCH or PUSCH. FIG. 2 shows the uplink subframe configuration that includes PUSCH and PUCCH.
If A/N is transmitted on the PUCCH, there are situations to be handled differently. For example, if the A/N transmission overlaps with the feedback of CSI (channel state information) periodically transmitted on the uplink, the PUCCH format 2a/2b is used. On the downlink, if carrier aggregation, in which transmission is performed using a plurality of carriers that are bundled together, is set to ON, and also the number of carriers is at least three, PUCCH format 3 is used. However, regardless of whether carrier aggregation is OFF or ON, if the number of carriers is two or fewer and there is no control information other than A/N and other than an uplink scheduling request, even if the number of carriers does not exceed two, PUCCH format 1a/1b are used. In considering that downlink data is transmitted more frequently than uplink data, and also considering that the period of CSI feedback is not more frequent than the period of downlink data assignment, A/N is most often transmitted by PUCCH format 1a/1b. The following description will focus on the PUCCH format 1a/1b. 
FIG. 3 shows the slot configuration of the PUCCH format 1a/1b. The A/N signals transmitted by a plurality of terminals are distributed by the Walsh sequence having a length-4 sequence and a DFT (discrete Fourier transform) sequence having a length-3 sequence and are code multiplexed and received at the base station. In FIG. 3, (W0, W1, W2, W3) and (F0, F1, F2) represent, respectively, the above-noted Walsh sequence and DFT sequence. At the terminal, a signal representing either ACK or NACK first undergoes primary spreading to frequency components corresponding to 1 SC-FDMA symbols by a ZAC (zero auto-correlation) sequence (with a subcarrier having a length-12 sequence) in the frequency domain. That is, a ZAC sequence having a sequence length of 12 is multiplied by an A/N signal component represented by a complex number. Then, the A/N signal after primary spreading and the ZAC sequence as a reference signal undergo secondary spreading by a Walsh sequence (W0 to W3 of a length-4 sequence, also called a Walsh code sequence) and a DFT sequence (F0 to F2 of a length-3 sequence). That is, each component of a signal having a length-12 sequence (an A/N signal after primary spreading or a ZAC sequence (reference signal sequence)) is multiplied by each component of an orthogonal sequence (for example, a Walsh sequence or a DFT sequence). Additionally, the signal after secondary spreading is converted by an IFFT (inverse fast Fourier transform) to a length-12 sequence (subcarrier) signal in the time domain. Then, a CP (cyclic prefix) is added to each signal after the IFFT, thereby forming a one-slot signal made up of seven SC-FDMA symbols.
A/N signals from different terminals having different cyclic shift indexes are spread using a ZAC sequence corresponding to different cyclic shift indexes and an orthogonal code sequence corresponding to different orthogonal cover indexes (OC indexes). The orthogonal code sequence is a set of a Walsh sequence and a DFT sequence. The orthogonal code sequence is also called a block-wise spreading code sequence. Therefore, by using the conventional despreading and correlation processing, the base station can demultiplex the plurality of A/N signals that have been code multiplexed and cyclic shift multiplexed. Because there is a limit to the number of A/N signals that can be code multiplexed and cyclic shift multiplexed per frequency resource block (RB), if the number of terminals becomes large, frequency multiplexing is performed using different RBs. In the following, the code-RB resource in which A/N is transmitted will be called the A/N resource. The A/N resource number is determined by the number of the RB in which A/N is transmitted, and the code number and cyclic shift value in the RB. Because multiplexing by cyclic shifting of the ZAC sequence can be treated as a type of code multiplexing, there will be cases in which orthogonal code and cyclic shift will be collectively called code in the following description.
In LTE, in order to reduce interference from other cells on the PUCCH, the ZAC sequence to be used is determined based on the cell ID. Because the mutual correlation between different ZAC sequences is small, by using different ZAC sequences between different cells, the interference can be reduced. Also, in the same manner, sequence hopping and cyclic shift hopping based on the cell ID has been introduced. With this hopping, shifting is done cyclically in units of SC-FDMA symbols, using a cyclic shift hopping pattern determined based on the cell ID, while mutual correlation on the cyclic shift axis and orthogonal code axis are maintained. Doing this enables randomization of combinations of A/N signals that are strongly interfered by other cells, while the mutual orthogonal relationship between A/N signals are maintained within a cell, and also enables removal of continuous strong interference to only some of the terminals from other cells.
In the description to follow, the description will be of the case in which a ZAC sequence is used for primary spreading, and a block-wise spreading code sequence is used for secondary spreading. However, for the primary spreading, rather than using a ZAC sequence, sequences that are mutually separable by mutually different cyclic shift values may be used. For example, a GCL (Generalized Chirp like) sequence, a CAZAC (constant amplitude zero auto correlation) sequence, a ZC (Zadoff-Chu) sequence, a PN sequence such as an M sequence or an orthogonal Gold code sequence, or a computer-generated random sequence having sharp auto correlation characteristics may be used for the primary spreading. As long as the sequence can be treated as being mutually orthogonal or substantially mutually orthogonal, any sequence can be used as a block-wise spreading code sequence for the secondary spreading. For example, a Walsh sequence or a Fourier sequence or the like can be used as a block-wise spreading code sequence for the secondary spreading.
In LTE, as a method of allocating different A/N resources to different terminals, allocation is used that is based on control information mapping results of the PDCCH. That is, using the fact that PDCCH control information is not mapped onto the same resources between a plurality of terminals, a one-to-one correspondence is established between the PDCCH resources and the PUCCH formats 1a/1b A/N resources (hereinafter described simply as A/N resources). This will be described below in detail.
PDCCH is made up of one or more L1/L2 CCHs (L1/L2 control channels). Each L1/L2 CCH is made up of one or more CCEs (control channel elements). That is, a CCE is the basic unit of mapping control information onto a PDCCH. Also, when one L1/L2 CCH is made up of a plurality (2, 4, or 8) of CCEs, a plurality of continuous CCEs with a CCE having an even-numbered index as the origin is allocated to that L1/L2 CCH. The base station, in accordance with the number of CCEs necessary for indication of control information to the subject terminal to which resources are to be allocated, allocates an L1/L2 CCH to the terminal to which the resources are to be allocated. The base station then maps the control information onto the physical resources corresponding to the CCE of that L1/L2 CCH. In this case, there is a one-to-one correspondence between each CCE and A/N resource. Therefore, a terminal that has received an L1/L2 CCH identifies the A/N resources corresponding to the CCEs making up that L1/L2CCH, and uses those resources (that is, codes and frequencies) to transmit the A/N signal to the base station. However, in the case of the L1/L2CCH occupying a plurality of continuous CCEs, the terminal uses an A/N resource corresponding to the CCE having the smallest index of a plurality of PUCCH constituent resources corresponding to a plurality of CCEs (that is, the A/N resource that has been associated with the CCE having a CCE index that is even number) to transmit the A/N signal to the base station. Specifically, the A/N resource number nPUCCH is established by the following equation (Non-Patent Literature 3).
[1]nPUCCH=N+nCCE  (Equation 1)
In this case, the above-noted A/N resource number nPUCCH is the above-described A/N resource number. N represents the A/N resource offset value given in common within the cell, and nCCE represents the number of the CCE having the smallest index onto which the PDCCH is mapped. According to Equation 1, it can be seen that an A/N resource within a certain range can be used in accordance with the range that can be taken by nCCE. In the following, the A/N resource that determines the resources dependent upon the control information scheduling of the PDCCH in this manner will be noted as D-A/N (dynamic A/N (dynamic ACK/NACK)).
As described above, the A/N resources include frequency resources in addition to code resources. Because the PUCCH and the PUSCH use the same frequency band in the uplink, there is a tradeoff between the region of the PUCCH that includes the D-A/N and the bandwidth of the PUSCH.